In enterprises and systems, such as computer networks, there is a need to provide for management of operations, in which the operations of the enterprise are detected and system changes are made to react to operational issues. In a conventional enterprise management system, the management process is a bottom-up approach in a disjointed system, where the management “plane”, in which data is obtained for analysis, is disconnected from the control “plane”, in which policies are developed and implemented.
Conventional systems may include different but related management and control processes. For example, in a FCAPS model of network management there are five parallel management components, generally called the fault-management component (F), the configuration management component (C), the accounting management component (A), the performance management component (P), and the security management component (S). However, conventional systems do not integrate these functions, but rather provide separate processes. For example, a network performance management may be provided by a system completely separated and different from the system that is managing network security, although network performance and security are two tightly-coupled functions that should be managed in close correlation.
Further, modern computer networks are increasingly diverse. A computer network often includes devices and systems that operate using various protocols and that are produced by various different manufacturers. The combination of diverse network elements and disconnected processes can create an inefficient management system that is not capable of quickly and effectively reacting to changes in the network.